GPT Migrator
APIs, integrations and technical migration

OpenAI file search: rebuilding GPT knowledge with vector stores

Last checked October 18, 2026
Short answerOpenAI's file search is a hosted tool in the Responses API that retrieves from files you uploaded to a vector store using semantic and keyword search. It supports file types including PDF, DOCX, PPTX, TXT, MD, JSON and HTML.

From OpenAI's developer documentation, checked October 18, 2026.

The flow

  1. Create a vector store.
  2. Upload your files with the Files API and add them to the store.
  3. Pass the vector store ID in the file_search tool on your Responses API call.

OpenAI says the Responses API is recommended for new projects and that the Assistants API was sunset on August 26, 2026; see rebuilding with the OpenAI API. Pricing for storage and calls is on OpenAI's pricing page, not the guide.

When to use it

It saves you from building your own retrieval if you are already building on OpenAI. Otherwise see how retrieval works and moving a knowledge base.

Not sure it is simple enough to do yourself?

Tell us what your GPT does, whether it uses Actions and how many files it has. We will say what will break and what a rebuild involves. The audit is free.

Request a free audit

Related guides

Sources

Pages describing other companies' products reflect their public documentation on the date shown and can go out of date.